September, 2016

Bay area bike share

  • One year (9/2014 - 9/2015).
  • Five cites (Mountain View, Palo Alto, Redwood City, San Jose, San Francisco).
  • Total of seventy stations.
  • 354152 bike trips.

How it works

  • 24 hour or 3 day passes can be bought at a station kiosk.
  • 1 year subscriptions available online. 88 per cent of trips are by subscribers.
  • Bikes can be hired and dropped off at any station.
  • Users can make as many 30 minute trips as they like. Longer trips incur late fees. ($4 for 30-60 minute trip, $7 each additional 30 minutes.)

People love it!

  • On average over 1200 trips per day are made during the week.
  • Over 360 per day are made on weekends.
  • Usage drops around January. It drops by 30 per cent on rainy days.

Holidays

  • Usage drops on holidays. Note Independence day fell on a saturday so was observed on friday the 3rd.

Trip length

  • The average trip length is 17.43 minutes, 1.35 km.
  • 95.5% of trips are made within the 30 minute time limit.

Long trips

  • The longest trip made lasted 200 days! ($67153 late fees…)
  • The largest distance between stations was 68 km, from San Jose to San Francisco, a trip of more than 8 hours:

Trip time: Weekday

-During the week trips are commonly made during commuting hours.

Trip time: Weekend

-During the weekend trips are taken more… leisurely.

Typical weekday (Use controls to zoom.)

Expansion!

  • The bike share is expanding ten-fold to 7000 bikes with new stations.
  • Users can submit suggestions for locations online. Over 5000 so far:

  • People are clearly enthusiastic, great! But how to choose?

Usage data

  • We'll look at the data, analyse usage patterns and suggest locations.
  • The data is available here.

Busiest stations

  • We will use a graph ranking algorithm to find the busiest stations.
  • Bike stations are nodes, links between nodes represent trips.
  • Weights on the links give the number of trips.

Eigenvalue centrality

  • Takes into account the number of links made by a node but also their quality (do they connect to busy stations?).
  • The higher the centrality value the more important the station.
  • We calculated eigenvalue centrality on a city by city basis.

Mountain View

Palo Alto

Redwood City

San Francisco

San Jose

Number 1 in each city:

  • Mountain View: Mountain View Caltrain station.
  • Palo Alto: Palo Alto Caltrain station.
  • Redwood City: Redwood City Caltrain station.
  • San Francisco: San Francisco Caltrain (Townsend at 4th).
  • San Jose: San Jose Diridon Caltrain station.

Caltrain stations

  • We see that the busiest locations in each city are at the Caltrain stations.

San Francisco Caltrain

  • 95 per cent of usage here is long term subscribers. There are two stations.
  • Bikes are picked up in the morning, dropped off in the evening by commuters.

Zip codes of commuters to SF Caltrain

  • People travel from all over (780 zip codes). Top 80 account for 92 per cent:

Expansion

  • Part of the planned expansion of the system is to have stations in Oakland, Emeryville and Berkeley.
  • Integrating with the Bay area rapid transit stations and other public transport will encourage commuters to use the bikes.

Expansion

  • There are already 35910 trips made from users who live here.

SF open data

  • The City and County of San Francisco has measured bicycle usage.
  • They have records going back to 2006 at various locations around the city.
  • The data is available here.
  • We'll take a look to see if we can get an idea of where bicycles are being used.

Bike usage has generally increased

  • Each line represents a location throughout the city.

Visualise where cycling is popular

High usage

  • Cycling common in Marina District, Lower Haight and Mission District.
  • The Golden gate bridge is popular too. According to San Francisco wikitravel: A very popular ride for visitors to San Francisco is the ride across the Golden Gate Bridge to Sausalito.

Recreation

  • Also note recreational areas like the park are also popular.
  • TripAdvisor will give us the tourist hot spots:

  • We see many are not catered for by BABS. Expansion into these areas may encourage more leisure, weekend use.

Imbalance

  • If more cycles are hired than deposited from a station then the station will require manual re-stocking.
  • If less cycles are hired than deposited the station fills up and depositing a bike can be frustrating.

Availability at S.F. Caltrain (Townsend at 4th)

  • We see that during peak hours it can be a problem.

Daily imbalance

  • We calculate the daily average deposits minus hires at each station:

Imbalance in San Francisco

Increasing the size of these stations will help alleviate system disruption:

Station Mean daily imbalance
San Francisco Caltrain (Townsend at 4th) 23
Grant Avenue at Columbus Avenue -11
2nd at Folsom -9
Market at Sansome 7
Powell at Post (Union Square) -7

Station Mean daily imbalance
Beale at Market -6
Market at 10th -5
2nd at Townsend 4
Temporary Transbay Terminal (Howard at Beale) -4
Townsend at 7th 4

Conclusion and recommendations

  • The BABS system is popular with commuters.
  • Peak usage is during typical commuting hours.
  • Usage drops on rainy days and holidays.
  • The busiest stations are located at transport hubs like Caltrain stations.
  • Expansion to new cities should be along existing public transport infrastructure.
  • Within San Francisco cycling is already popular in Marina District, along the 101, in Lower Haight and Mission District. Bike stations here will be welcome.
  • Increasing the size of the most imbalanced stations will help mitigate any disruption to users caused by lack of space or bikes.

Conclusion and recommendations

  • Recreational areas like the parks are currently popular with cyclists but not catered for by BABS. Bike stations in these locations will promote the BABS as a leisure activity.
  • Promotion to tourists as a travel option would help too: There is currently no mention of BABS on San Francisco wikitravel or Bay area wikitravel.
  • There is some confusion about how the system works, some users don't seem to understand the 30 minute rule (or don't care!).

Happy cycling!